[RFC][PATCH 1/2] Track in-kernel when we expect checkpoint/restart to work

[Dave Hansen]

On Fri, 2008-10-10 at 10:37 +0200, Daniel Lezcano wrote:
> For example, you create a socket, the process becomes uncheckpointable, 
> you close (via sys_close) the socket, you have to track this close to be 
> related to the socket which made the process uncheckpointable in order 
> to make the operation reversible.

Challenging, yes.  Not quite a nightmare, though.

It's basically the same problem we have with r/o bind mounts individual
users in the kernel need to check some global state.  There are
temporary (like the duration of the syscall) and more chronic (like your
sockets) users.  The hard part is making sure that everything gets
properly balanced with the chronic ones.  But, we did manage to
accomplish that with the r/o bind mounts.  I haven't seen a bug in
*hours*! ;)

In any case, we can worry about that later.  

-- Dave